Smartwatch Occlusion Image Dataset
V1.0Latest Update: | 2025-10-16 | Samples: | 5000 records |
Samples: | 5000 records | ||
File Size: | 1.2G | Format: | JPG/PNG/JSON |
Format: | JPG/PNG/JSON | ||
Data Domain: | Image | Holder: |  MOBIUSI INC |
Holder: | MOBIUSI INC | ||
Industry Scope: | Gesture Recognition | User Interaction Analysis | Product Testing | ||
Applications: | Object Detection | Image Classification | ||
Brief Introduction
The retail e-commerce industry is facing challenges with accurate gesture recognition and user interaction analysis due to obstructions such as hand gestures, wristbands, reflections, or blurred watch faces. Current solutions often lack robustness in varying lighting conditions and occlusion types, leading to poor user experience. This dataset aims to address these specific technical issues by providing a comprehensive collection of images highlighting various occlusion scenarios. The data was collected using high-resolution cameras in controlled environments, ensuring clarity and consistency. Quality control measures include multi-round annotations, consistency checks, and expert reviews to verify accuracy. The dataset is stored in JPG format, organized by occlusion type and timestamp for easy access and analysis.
Sample Examples
| Image | File Name | Resolution | Image Brightness | Image Contrast | Degree of Occlusion | Gesture Type | Watch Visibility | Lighting Conditions | Image Blur | Background Complexity |
|---|---|---|---|---|---|---|---|---|---|---|
 | ce278d4a37e6480ee0be79583a8c374d.jpg | 1224*1632 | Medium | Moderate | None | None | Fully visible | Indoor | Clear | Normal |
 | 757bf9ca703437f7a63ff3b7649c2abf.jpg | 1224*1632 | Moderate | Medium | Unobstructed | Wrist extended | Over 90% | Indoor | Clear | Complex |
 | 09320243027d79aec44849f5e0ce38a0.jpg | 1227*1629 | Medium | Normal | Unobstructed | None | Fully visible | Indoor | Clear | Simple |
 | c00dcc13242ab45ea904e2040bf8ed70.png | 1347*1500 | High | Medium | Unobstructed | No gesture | Fully visible | Indoor | Clear | Simple |
 | 4fecc97c24e3cf81b0b22ab023787495.jpg | 1224*1632 | Medium | Low | Partially obstructed | Hand holding position | Mostly visible | Indoor | Slightly blurred | Normal |
 | 233ca1416cc62ee32f0fb42b09c92f29.png | 1500*1498 | Medium | High | Unobstructed | None | Fully visible | Indoor | Clear | Simple |
 | d1fb7621b48a6c24fc0394bcedaea380.jpg | 961*2080 | Low | Low | Partially obstructed | Holding | About 50% | Indoor | Blurred | Normal |
 | 5369068867c578d2c216b6d92b0c9810.jpg | 1224*1632 | Medium | Moderate | Unobstructed | No gesture | Fully visible | Indoor | Quite clear | Normal |
 | b57c1f30a6b3c80d68b786eb3c12efca.jpg | 1224*1632 | Overall image brightness is moderate, with good lighting conditions | Good contrast, visually clear | Unobstructed, smartwatch fully visible | No gesture, as no hand is present | Smartwatch visibility area is 100% | Indoor lighting | Clear image, no obvious blurriness | Simple background, few distractions |
 | 2c4dd280f536d92f9f4c9150f24cdff4.jpg | 1224*1632 | Medium | Medium | Unobstructed | No gesture | Fully visible | Indoor | Clear | Simple |
Data Structure
| Field | Type | Description |
|---|---|---|
| file_name | string | File name |
| quality | string | Resolution |
| image_brightness | float | The average brightness value of the image, used to evaluate the lighting conditions. |
| image_contrast | float | The contrast value of the image, reflecting the visual clarity of the image. |
| occlusion_level | string | The extent to which the smartwatch is obscured in the image, such as unobstructed, partially obstructed, or completely obstructed. |
| hand_gesture | string | Gesture recognition tags in the image, identifying the posture and movement of the hand. |
| watch_visibility | float | The proportion of the visible area of the smartwatch in the image, used to assess the watch's visibility. |
| lighting_condition | string | The lighting environment at the time the image was taken, such as daytime, nighttime, indoor, or outdoor. |
| image_blurriness | float | The degree of blur in the image, used to assess the clarity of the image. |
| background_clutter | string | The complexity of the image background, such as simple, ordinary, or complex. |
Compliance Statement
| Item | Content |
|---|---|
| Authorization Type | CC-BY-NC-SA 4.0 (Attribution–NonCommercial–ShareAlike) |
| Commercial Use | Requires exclusive subscription or authorization contract (monthly or per-invocation charging) |
| Privacy and Anonymization | No PII, no real company names, simulated scenarios follow industry standards |
| Compliance System | Compliant with China’s Data Security Law / EU GDPR / supports enterprise data access logs |
Can't find the data you need?
Post a request and let data providers reach out to you.